IB9CJ Asset Pricing (PhD)
Data: From Kenneth French’s web site,
mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html,
download monthly data on …
• Fama-French (“FF”) factors,
• Momentum factor,
• 25 (5 × 5) portfolios formed on size and book-to-market (“BM”), and
• 30 Industry portfolios.
Separate the 100 most recent observations in the time series and designate these for “out-of-sample” analysis. For the “in-sample” data, select observations from 1980:01 until the month preceding the beginning of the out-of-sample period.
To conduct the numerical analysis, you have to use matlab.
Q1: Efficient Portfolios.
1.1. Using in-sample data, estimate the VCV matrix for the industry portfolios. Based on this, construct the efficient frontier. Explicitly describe the two portfolios that “span” the frontier (as in the ‘two-fund separation’ theorem).
1.2. From the in-sample factor data, extract an appropriate value to be used as the risk-free rate out-of-sample, and explain your choice. With this, construct the “tangency portfolio”, and discuss its (in-sample) properties.
1.3. From the in-sample factor data, extract the returns on the “market portfolio”. Estimate its expected return and volatility. Is the market portfolio efficient? Discuss your findings, in particular in the context of the CAPM.
1.4. Produce a plot, showing the in-sample characteristics of the assets, the frontier, and the all of the portfolios constructed in 1.1 – 1.3 above.
1.5. Track the returns of the portfolios considered above through the out-of-sample period. Using measures of your choosing (explain the rationale of your choice), assess and discuss the performance of these portfolios out-of-sample.
Q2: Factor Models
Note: This question may appear “longer” than Q1. However, you will find that some of the methodology (and even results) that you will have developed for Q1 can be “re-cycled” here.
2.1. Using in-sample data, estimate the “betas” of the size-BM portfolios…